In an exclusive interview during the Merchant Payments Ecosystem 2024, Brittany Allen, a trust and safety architect at Sift, shared insights on the evolving landscape of fraud, particularly emphasizing the concerns surrounding misinformation propagated through social media and AI-driven technologies.
Allen, who has held pivotal roles in fraud prevention at prominent companies like Etsy, Airbnb, and 1stDibs, brings extensive expertise to the discourse on financial fraud. She illustrates the myriad ways fraud is perpetrated in the digital marketplace, ranging from e-commerce scams to luxury fraud. Notably, she recounted her team’s close call with Anna Delvey, the infamous con artist who became widely recognized through a Netflix portrayal.
A significant trend Allen identified is the rise of “refund” or “return fraud.” She stresses that educating consumers is vital for both e-commerce platforms and financial institutions, particularly in an era where scammers are increasingly sophisticated. The emergence of “fraud influencers”—individuals who promote fraudulent activities in an almost business-like fashion—has made it essential for consumers to be wary of scams.
“In the past, fraudsters might have used dark web marketplaces to acquire stolen credit cards for illicit purchases. Now, they can leverage platforms like Telegram to reach a broader audience, often providing mentorship and instructional content on committing refund fraud,” Allen explains.
She points out that the surface web is also witnessing an uptick in fraud-related discussions, aiding the recruitment of new scammers through deceptive channels. A typical scenario involves unsuspecting individuals stumbling onto social media platforms like TikTok, where they might inadvertently encounter content promoting fraudulent schemes. “This further perpetuates a cycle, leading users into a rabbit hole of fraud-related content and recruitment invitations via links to private channels,” she adds.
Allen highlights the complexities businesses face in detecting these fraudulent activities. For instance, on platforms like TikTok, scammers can embed illicit content in subtle ways—through captions, video banners, and text boxes—making it increasingly challenging for companies to identify and counteract fraudulent behavior.
Transitioning to the discussion of AI and deepfake technology, Allen notes that while the impersonation of close acquaintances for scams is less likely due to the effort required, the fraudulent use of celebrity likenesses poses a significant threat—particularly concerning solicitations for charitable donations.
She emphasizes the rapidity with which misinformation can spread, especially in the context of upcoming electoral events. “This will be the first major U.S. federal election where access to AI and deepfake tools is widespread. We must be prepared for the likelihood of manipulated content featuring political candidates, which may stem from various entities, including foreign governments looking to interfere,” Allen warns.
An example of misinformation’s impact can be seen in the recent controversy surrounding a digitally altered image of Princess Kate and her family. Allen urges individuals to verify information sources diligently, noting that initial perceptions can often be misleading without thorough investigation.
On the subject of safeguarding against payment fraud, Allen advises financial institutions and merchants to collaborate more effectively. “Merchants have access to comprehensive data about customer interactions on their platforms, while banks possess insights into broader spending patterns. Bridging the information gap between these entities is crucial for enhancing risk assessment,” she emphasizes.
She also points out ongoing discussions regarding the responsibilities of financial institutions in combatting fraud and the regulatory frameworks still being established to guide their actions in preventing fraudulent transactions.
Addressing the role of AI and machine learning in fraud detection, Allen expresses concerns about ensuring that datasets are unbiased and do not lead to discriminatory outcomes. “While the vision of a fraud-free future is appealing, it may be unrealistic. Striking a balance between managing fraud risk and serving legitimate customers is essential,” she asserts.
Looking forward, Allen envisions a landscape where fraud detection systems are not only effective but also transparent, allowing merchants to grasp the reasoning behind decisions made by AI algorithms. Such clarity would empower businesses to understand and potentially influence future fraud prevention measures.